The PCB is mounted in a NEMA 4 enclosure which is outside in a open to=20
the weather (blowing snow, but not direct rain) and dew, not much dust=20
is possible, but appearance of other items in the enclosure does not=20
suggest much problem due to the environment. You are suggesting changing=20
the opt-isolator every year or 2 as good maintenance.

I just noticed on a datasheet, 2 items of interest:
1: The collector > 12 VDC resistor which I show, and is as 100K ohms,=20
Vishay suggests 100 ohm. This is the connection to the Mosfet "G"
2: I have the Opto out from pin 5 "C" instead of pin 6 "B". But it is=20
suggested for immunity from high electromagnetic interference something=20
about no base connection. The opto control part has been working good.

On 1/8/2012 4:54 PM, Bob Blick wrote:
> I guess something is causing the mosfet to get a couple of volts of
> gate, and it sits warming until it goes pop.
>
> Perhaps you don't have enough drive on the optocoupler (the transfer
> ratio goes down by half over a few years).
>
> Or perhaps there is moisture/dew/dust?
>
> Bob
>
> On Sun, Jan 8, 2012, at 04:36 PM, Carl Denk wrote:
>   =20
>> I have a TIL157 NPN Darlington photo transistor controlling an IRF9530
>> Power Mosfet, that is controlling a 1" cube auto relay with typically
>> 133 ma coil. The circuit may be powered up for an extended period of
>> time, but generally less than 20 minutes. Here's a brief description: I
>> do have a small JPG of the circuit, just send me a private to attach to.
>>
>>     =20
>>> The TIL157
>>>      Emitter>  ground
>>>      Base>  N.C.
>>>      Colllector>  Mosfet G, with .01mfd to ground, and 100k ohm to + 12=
VDC  (Note this a a lead/acid batter and 14 VDC common)
>>>
>>> The Mosfet
>>>      S>  12 VDC
>>>      G>   From the TIL157
>>>      D>  to the relay coil with surge diode across the coil. Coil ~ 133=
ma
>>>
>>>
>>>       =20
>>
>> The circuit has been in use for 6 years, and this is the 2nd time I
>> needed to replace the Mosfet. There has been a rubber (??) insulator
>> with a machine screw to tie the mosfet down to the PCB as heat sink.
>> This time the PCB was black/brown color in the mosfet area - extreme
>> heat. The PCB is entirely Thru-the -hole.
>>
>> I would be happy if someone could review things and comment. I would
>> desire to keep the PCB's, cost is little priority since this is one
>> board with several spares.
>>     =20
>   =20
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.